Design of a simple workpiece size measurement system based on monocular vision

A lightweight non-contact size measurement system is proposed in this paper, based on embedded image acquisition and MATLAB image processing, enabling rapid and accurate workpiece size measurements. In the hardware configuration, STM3F407ZGT6 is adopted as the main controller to operate the OV2640 camera for image acquisition, while the captured images are displayed in real-time on an liquid crystal display (LCD) screen. The software component is developed on MATLAB platform, realizing essential functions including system calibration, distortion correction, image denoising, and edge detection, thereby achieving accurate calculation of the distance between any two designated points in the image. Experimental results indicate that the system has good measurement repeatability, with standard uncertainty of 0.023-0.030 mm and relative error of 0.08%-0.43% in sample measurements. Compared to conventional contact-based measurement approaches, the proposed system has the characteristics of non-contact operation and operational simplicity, rendering it applicable in industrial settings where physical contact is not feasible or where efficient measurement is necessary.